Research on News Video Multi-topic Extraction and Summarization

Di Li , Hua Huo

Comparing to other videos, news video contains more important information. Processing a news video with semantic extraction, abstract and annotate its sub-theme can facilitate our operations such as reading and retrieval, make the news video more easy to management and application. In this paper, we use news video text area positioning and extraction technology to obtain text corpus, put these sentences in Chinese word segmentation and syntactic analysis system, then computation sentence weighting and ranking them, calculate the similarity with the sentences. Finally, conduct a study about multi-topic extraction and summarization. The experimental result show this method with high pertinence which can better achieve the targets.